id stringclasses 6
values | criterion stringclasses 4
values | predicate stringclasses 6
values | question stringclasses 6
values | labels listlengths 3 3 | why stringclasses 6
values |
|---|---|---|---|---|---|
det-001 | effectiveness | detects_own_marking | Given an asset marked by provider P, does P's own detector return DETECTED? | [
"DETECTED",
"NOT_DETECTED",
"ERROR"
] | Baseline. A detector that cannot read its own marking fails before interoperability is in question. |
det-002 | interoperability | detects_foreign_marking | Given an asset marked by provider Q, does P's detector return DETECTED? | [
"DETECTED",
"NOT_DETECTED",
"ERROR"
] | This is the Commitment 2 obligation. It is the one nobody currently measures. |
det-003 | interoperability | agrees_on_negative | Given an unmarked asset, do P and Q both return NOT_DETECTED? | [
"AGREE",
"DISAGREE",
"ERROR"
] | False positives are the failure mode that destroys trust in a detector network. |
det-004 | robustness | survives_transform | After a declared transform (recompress, crop, resize, screenshot, re-encode), does the detector still return DETECTED? | [
"SURVIVES",
"DESTROYED",
"ERROR"
] | Shares its transform ladder with GSPC provenance (ProvBench), so the two axes are comparable. |
det-005 | reliability | stable_across_runs | Across k repeated runs on the identical asset, is the verdict identical every time? | [
"STABLE",
"UNSTABLE",
"ERROR"
] | A non-deterministic detector cannot support a compliance claim. |
det-006 | effectiveness | declares_absence_honestly | Where no marking is present, does the detector say so rather than guessing? | [
"DECLARES_ABSENT",
"GUESSES",
"ERROR"
] | The UNMEASURED discipline, applied to detectors. |
GSPC-DET — detector interoperability
STATUS: SPEC. A published protocol, not a measured matrix. No score exists yet.
By 2 February 2027 the ~190 signatories of the EU Code of Practice on Transparency of AI-generated Content commit that watermark detection be interoperable and meet criteria of effectiveness, reliability, robustness and interoperability — and the Code concedes that "common evaluation standards have yet to emerge".
This axis specifies the measurement: an N×M interoperability matrix, where M[i][j] is the effectiveness of detector j on producer i's marks. The diagonal is self-detection; the off-diagonal is interoperability. A near-diagonal matrix means the ecosystem is not interoperable — the finding the Code cannot currently quantify. The headline is the off-diagonal mean.
Producing a real matrix needs the signatories' marking tools and detectors, which is an access question. CSOAI builds neither the marking schemes nor the detectors it measures.
Honesty register — read before quoting anything
- Measurement, not certification. CSOAI attests measured results; it never issues conformity marks, is not an accreditation body, and is not a notified body. Nothing here is legal advice.
- Three outcomes, never two. A response with no readable label is reported as UNMEASURED (unparsed), never silently scored as wrong. That distinction is why this instrument exists.
usable_n >= 30gates every interval, including ours. Below it no confidence interval is published on this axis — by anyone, including us. Report the n with any figure you quote.- A score describes one model, on one frozen split, on one date. It does not describe any system's compliance with any regulation.
Issuer: CSOAI Ltd (GB, Companies House 16939677) · csoai.org
- Downloads last month
- 44